GAF proved to be a reliable and, within the limits of the indicators used, a valid measure of psychiatric disturbance in our sample of the severely mentally ill. Differences in relationships between the three GAF scores and medication/support needs indicate the usefulness of obtaining all three scores for monitoring levels and type of psychiatric disturbance in this population.

Two hypotheses were tested concerning the nature of the cognitive dysfunction in schizophrenia: (a) that there is a broadening of selective attention; and (b) that there is an impairment in associational learning. RDC-diagnosed acute and chronic schizophrenics and normal controls carried out a choice reaction time (RT) task in which conflict between the correct response to a target (a letter in the centre of a computer screen) and that cued by simultaneously presented flankers (two letters either side of the target) increased RT. For 80 ('valid') trials, flankers and targets were consistent in the response cued (pressing a button with either left or right hand); on 8 ('invalid') trials they conflicted. On invalid trials there was a slowing of RT, and an increase of errors for left-hand responses. Chronic schizophrenics showed the same reactions to cue validity as normal controls, both groups differing significantly from acute schizophrenics. For the latter, the RT data supported hypothesis (b), but the error rates appeared to support hypothesis (a).

Recently, research on mental health conditions using public online data, including Reddit, has surged in NLP and health research but has not reported user characteristics, which are important to judge generalisability of findings. This paper shows how existing NLP methods can yield information on clinical, demographic, and identity characteristics of almost 20K Reddit users who self-report a bipolar disorder diagnosis. This population consists of slightly more feminine-than masculinegendered mainly young or middle-aged USbased adults who often report additional mental health diagnoses, which is compared with general Reddit statistics and epidemiological studies. Additionally, this paper carefully evaluates all methods and discusses ethical issues.

The Kamin blocking effect consists in impaired learning of an association between a conditioned stimulus (CS2) and an unconditioned stimulus (UCS) if CS2 is presented simultaneously with a different CS (CS1) already associated with the UCS. It is well established with animal but not human subjects. In the two experiments presented here, the effect was clearly demonstrated with UCS = a computer-presented yellow square, CS1 = a blue square and CS2 = a noise (Expt 1) or two smaller white squares (Expt 2). The Kamin effect has, by some theorists, been attributed to switching of attention away from CS2, which conveys only redundant information. If so, the size of the effect should co-vary with indices of selective attention commonly used with human subjects. As predicted, the blocking effect was found to be smaller in subjects who displayed a high degree of incidental learning in either of two tasks in which intentional vs. incidental learning corresponded to (1) words vs. word position, or (2) a target initial word letter vs. non-target initial letters. We report elsewhere that the blocking effect is absent in acute schizophrenics. In the present experiments, however, there was no systematic relation between any of four measures of psychoticism, or intelligence, and either the blocking effect or incidental learning.

Recently, research on mental health conditions using public online data, including Reddit, has surged in NLP and health research but has not reported user characteristics, which are important to judge generalisability of findings. This paper shows how existing NLP methods can yield information on clinical, demographic, and identity characteristics of almost 20K Reddit users who self-report a bipolar disorder diagnosis. This population consists of slightly more feminine-than masculinegendered mainly young or middle-aged USbased adults who often report additional mental health diagnoses, which is compared with general Reddit statistics and epidemiological studies. Additionally, this paper carefully evaluates all methods and discusses ethical issues.